Measure Okr Grader

<!-- PM-Skills | https://github.com/product-on-purpose/pm-skills | Apache 2.0 -->

OKR Grader

An OKR Cycle Review is a backward-looking artifact that closes the loop on a completed OKR set. It scores each KR against its baseline and target, separates committed from aspirational interpretation, surfaces what evidence does and does not support, names what the team learned, and prepares input for next-cycle drafting. Done well, a cycle review protects the integrity of the OKR operating system by refusing to dress up missed commitments as aspirational stretch, refusing to celebrate effort over outcome, and refusing to let scoring carry weight it cannot bear.

This skill is an evidence interpreter, not an arithmetic engine. Its job is to read final KR values, compare them against the original OKR set's intent, and produce a review that names the learning honestly. It enforces the empirical scoring conventions drawn from Doerr (Measure What Matters), Wodtke (Radical Focus), Castro (committed vs aspirational interpretation), Grove (High Output Management), and the OKR community's accumulated practice on misuse failure modes. It pairs with foundation-okr-writer (which produced the OKR set being scored) and hands off the learnings produced here to the iterate skills that consume them.

When to Use

• The OKR cycle has ended (or you are scoring a partial-cycle close)
• You have final or interim KR values, baselines, and targets
• Stakeholders need a clear review with score, evidence, and learning
• The team is deciding what to continue, stop, change, or carry forward
• There is disagreement about whether a score is good or bad
• Evidence quality across KRs is uneven and needs to be made visible

When NOT to Use

• You are still drafting OKRs - use foundation-okr-writer
• You want a generic team retro - use iterate-retrospective
• You are reporting a single experiment result - use measure-experiment-results
• You need a stakeholder progress update without scoring - use foundation-stakeholder-update
• The OKR set was never agreed on or never tracked - scoring requires an authored set; backfill via foundation-okr-writer first
• You want to use scores to evaluate individuals - the skill refuses this

Instructions

When asked to score completed OKRs, follow these steps:

1. Validate scoring readiness Check inputs: original OKR set, cycle dates, final KR values (or interim values for partial-close), baselines, targets, evidence sources, and OKR types (committed | aspirational | learning | operational_health | compliance_or_safety). If a value is missing, mark it explicitly (not-yet-observable, not-instrumented, not-supplied); never fabricate. Refuse to grade KRs whose original definitions are missing entirely.

2. Classify each KR's type and indicator class The OKR type is one of committed | aspirational | learning | operational_health | compliance_or_safety (the five values produced by foundation-okr-writer). The indicator class is one of leading | lagging | guardrail | health | evidence_generation. Carry both forward from the original OKR set, or assign defaults if the original set did not specify. The OKR type determines the scoring convention: aspirational uses the 0.6 to 0.7 sweet spot; committed targets 1.0; compliance_or_safety is binary; operational_health is pass | fail | drift-within-tolerance against a threshold band; learning grades by validated or invalidated rather than by score. The indicator class adds independent rules that apply on top of the type's scoring (see Step 3).

3. Score each KR For each KR, compute or assign a score using the convention for its OKR type:

• aspirational KR: numeric score = (actual - baseline) / (target - baseline). Sweet spot is 0.6 to 0.7.
• committed KR: pass or fail against the target. Anything below 1.0 is a miss.
• compliance_or_safety KR: binary. Met or not met. No partial credit. No retroactive scope shrinkage when coverage is partial; mark as not-yet-fully-observable instead.
• operational_health KR: pass | fail | drift-within-tolerance against the threshold band.
• learning KR: validated, invalidated, partially-validated, or insufficient-evidence. No numeric score.

Then apply indicator-class rules independently of the OKR type:

• any KR with indicator class guardrail is reported as its own signal and is NEVER averaged into the primary objective score, regardless of its OKR type. A failed guardrail does not dilute a high primary KR score.

For each score, state the calculation or rationale and the evidence confidence (high | medium | low | unknown).

4. Interpret the objective score Avoid naive averaging when one KR is a guardrail, compliance threshold, or learning KR. Produce a qualitative read of the objective alongside any rough numeric average. State explicitly what the score does and does not mean.

5. Assess evidence quality For each KR, name the evidence's reliability and any caveats (instrumentation gaps, target shifts mid-cycle, cohort definition changes, measurement window mismatches, sample-size limitations). Recommend fixes for next cycle's measurement plan.

6. Review initiatives as bets For each initiative the team ran, name which KR it was expected to move, whether it shipped, what its apparent contribution was, and whether the evidence supports continuing, retiring, or reworking it. Use Castro's "initiatives are bets, not commitments" framing. Separate ship-status from KR-impact; an initiative that shipped on time but did not move its KR is not a partial win.

7. Synthesize learning Capture validated assumptions, invalidated assumptions, surprises, and decision implications. Distinguish between learnings about the customer or product (carry forward), learnings about team process (hand to iterate-retrospective), and learnings about measurement (hand to measure-instrumentation-spec or measure-dashboard-requirements).

8. Prepare next-cycle recommendations For each objective: continue, revise, retire, or escalate. Suggest candidate next-cycle OKRs or open questions for foundation-okr-writer. Hand-off measurement gaps to measure-dashboard-requirements or measure-instrumentation-spec. Hand-off assumption tests to define-hypothesis. Hand-off team-process work to iterate-retrospective. Hand-off organizational memory to iterate-lessons-log. Hand-off next-cycle drafting to foundation-okr-writer.

9. Surface risks in interpretation Make explicit any places the score could mislead a reader: forced numeric scores on KRs that are not yet observable, confounded initiative results, stakeholder framings that under-state evidence, single-cycle results that need a second cycle of confirmation.

10. Note the source of truth The artifact is a review document, not the canonical OKR system. Include a source_of_truth field pointing to the original OKR tracker.

11. Finalize for direct use Remove all skill instruction commentary from the final artifact. The final output should be reader-facing.

Constraint Rules (MUST / MUST NOT)

These rules are non-negotiable. The skill enforces them in every grading run.

• MUST NOT retroactively change baselines, targets, or KR definitions. If the team adjusted these mid-cycle, document the change explicitly and grade against both the original and adjusted versions.
• MUST NOT retroactively shrink the scope of a committed or compliance_or_safety KR to mark partial coverage as a pass. If the original commitment named 3 healthcare accounts and only 1 has been audited, the KR is not-yet-fully-observable. The 1-account result is a sub-signal, not the KR score.
• MUST NOT treat 0.7 as success for committed, compliance_or_safety, or operational_health KRs. Those target 1.0 (or the threshold band).
• MUST NOT average away a failed guardrail. A failed guardrail is a separate signal that does not get diluted by the primary KR's success.
• MUST NOT equate effort with impact. Initiatives that shipped on time but failed to move their KR are not partial wins.
• MUST NOT use OKR scores as individual performance ratings or compensation inputs. If the user requests this, refuse and explain the sandbagging and learning-suppression risks.
• MUST NOT punish honest stretch when aspirational intent was explicit and disclosed at OKR-writing time. A 0.6 aspirational score is the designed sweet spot.
• MUST NOT celebrate missed committed goals as ambitious failure. Committed misses are misses.
• MUST mark any not-yet-observable KR explicitly (e.g., a 90-day retention cohort whose window extends past cycle close). Forced numeric scores on not-yet-observable KRs are misleading.
• MUST include evidence confidence on every KR score (high | medium | low | unknown).
• MUST NOT become the canonical source of truth. Always include a source_of_truth pointer to the user's actual OKR tracker.

Scoring Rules

The skill applies these conventions to every cycle review. The convention follows the OKR type, not the team's preference at grading time. OKR type and indicator class are independent dimensions; type controls scoring, indicator class adds reporting rules.

OKR types determine the scoring convention:

• `aspirational`: numeric score on a 0 to 1 scale = (actual - baseline) / (target - baseline). Sweet spot is 0.6 to 0.7. Below 0.4 is a miss; above 0.8 over multiple cycles suggests sandbagged targets needing recalibration.
• `committed`: pass or fail against the target. Anything below 1.0 is a miss requiring postmortem. Do not soften with aspirational interpretation.
• `compliance_or_safety`: binary. Met or not met. No partial credit. No retroactive scope shrinkage. If the committed scope is only partially observable (some audits pending, some accounts deferred), mark the KR as not-yet-fully-observable; the observed subset is a sub-signal, not the KR score.
• `operational_health`: pass | fail | drift-within-tolerance against the threshold band.
• `learning`: validated | invalidated | partially-validated | insufficient-evidence. No numeric score.

Indicator class rules apply on top of the OKR type's scoring:

• indicator class `guardrail`: the KR is scored per its OKR type, and additionally is reported as its own signal, never averaged into the primary objective score. A failed guardrail does not dilute a high primary KR score, regardless of whether the guardrail itself is committed, aspirational, operational_health, or compliance_or_safety.

Special states:

• `not-yet-observable`: score deferred. Do not force a numeric score; mark interim signal and projected score with explicit confidence and the date the final score becomes available.
• `not-yet-fully-observable`: a committed or compliance_or_safety KR with partial coverage. Score the KR as deferred until full coverage is observable. Do NOT promote a sub-signal to a KR-level pass.

Anti-Patterns the Skill Detects

The skill scans for these and either flags or refuses:

• Retroactive target adjustment (we hit it because we changed the target) - document the change; grade against both definitions
• Retroactive scope shrinkage on a committed or compliance_or_safety KR (committed to 3 healthcare audits, 1 audit completed, scored as "pass on in-scope") - refuse and mark not-yet-fully-observable
• Average-the-guardrail-away (a failed guardrail dissolved into a high primary score) - separate the guardrail signal
• Aspirational-grading-of-committed (treating 0.7 as success on a committed KR) - refuse and explain
• Effort-equals-impact (initiative shipped, score did not move, scored as partial win) - separate ship-status from KR-impact
• Compensation coupling (using the score for performance reviews) - refuse and explain
• Missed-committed-as-stretch (we did not quite hit the contractual deadline but the team really tried) - refuse the framing
• Sandbagged target (consistently scoring above 0.85 on aspirational targets) - flag for next-cycle target recalibration
• Forced score on not-yet-observable (giving a numeric score to a KR whose 90-day window has not closed) - mark deferred
• Initiative-as-cause-without-evidence (claiming Initiative X drove KR Y when timing or instrumentation cannot support it) - separate apparent contribution from causal claim
• Hidden low-confidence (precise numeric scores with weak evidence) - surface confidence; do not let precision mask uncertainty
• Stakeholder narrative override (a leader's preferred framing taking precedence over the evidence) - the grader's read is independent of stakeholder framing
• Single-cycle confirmation (treating one cycle's signal as proof) - recommend a second cycle when the evidence is suggestive but not robust

Output Contract (v1.0.0)

• All required sections present in canonical order: Summary, Scorecard, Objective Interpretation, Evidence Quality, Initiative Review, Learning, Next-cycle Recommendations, Risks in Interpretation
• Every KR in the Scorecard includes: actual value (or not-yet-observable / not-yet-fully-observable marker), score using the type-appropriate convention, evidence confidence, interpretation
• aspirational KRs use the 0 to 1 numeric scale; committed KRs are pass or fail; compliance_or_safety KRs are binary; operational_health KRs are pass | fail | drift-within-tolerance; learning KRs use validated or invalidated language
• KRs with indicator class guardrail are surfaced separately and never averaged into the primary objective score, regardless of OKR type
• Partial-coverage on a committed or compliance_or_safety KR is marked not-yet-fully-observable, not pass-on-in-scope
• Source-of-truth note is present and points to a non-skill location
• Hand-off section names specific downstream skills for learnings, team-process work, assumption tests, and measurement gaps
• Markdown only output. No JSON.
• Measure phase classification: phase: measure in frontmatter; no classification: field

Quality Checklist

Before finalizing, verify:

• [ ] Every KR has a final value, an explicit not-yet-observable marker, or an explicit not-yet-fully-observable marker (for partial-coverage on committed or compliance_or_safety KRs)
• [ ] Every KR has an evidence confidence rating
• [ ] Every KR's score uses the convention for its OKR type from the canonical enum: committed | aspirational | learning | operational_health | compliance_or_safety
• [ ] guardrail is treated as indicator class, not as an OKR type
• [ ] KRs with indicator class guardrail are surfaced separately and never averaged into the primary score
• [ ] No retroactive target changes are silently absorbed
• [ ] No retroactive scope shrinkage on committed or compliance_or_safety KRs (partial coverage is not-yet-fully-observable, not pass-on-in-scope)
• [ ] No committed KR is graded as aspirational
• [ ] No effort-equals-impact framing on initiatives
• [ ] No compensation-coupled framing
• [ ] Risks-in-interpretation section names where the score could mislead a reader
• [ ] Hand-off section names specific downstream skills with rationale
• [ ] Source-of-truth note present
• [ ] Skill instruction commentary removed from final artifact
• [ ] Markdown only - no JSON output

Examples

See references/EXAMPLE.md for a completed cycle review in the storevine sample thread (Campaigns team, Q3 2026 close), demonstrating aspirational scoring with one KR not-yet-observable, a held guardrail, and a templates-as-retention-driver thesis invalidation. The companion foundation-okr-writer skill produces the OKR sets this skill scores; together they cover the full quarterly arc.

Scenario: scoring the Activation team's Q3 OKRs at cycle close

This is the INPUT brief for an output-quality eval. The skill arm and the control arm each receive everything below (and nothing else about how to do the work) and produce an OKR cycle-review artifact for it. Judges never see this header.

The OKR set being scored (as authored at the start of Q3)

Objective: "New teams reach value fast enough to stick and convert."

KR1 (aspirational): Increase signup-to-activation rate from 22% baseline to 35% target.

• Final value at cycle close: 30%.

KR2 (committed): Ship and instrument the new onboarding checklist so median time-to-first-dashboard is measurable for 100% of new teams by end of quarter.

• Final value: instrumentation shipped for ~80% of new teams (Android rollout slipped two weeks);

the other 20% are not yet measured.

KR3 (aspirational): Increase free-to-paid conversion within 30 days from 6% baseline to 9% target.

• Final value: the 30-day conversion window for September signups closes Oct 15, after cycle end.

As of cycle close, the August cohort is at 7%; September is not yet observable.

KR4 (guardrail): Keep new-user support tickets per 100 signups at or below the baseline of 14.

• Final value: support tickets rose to 19 per 100 signups during the checklist rollout.

Initiatives the team ran:

• Redesigned onboarding checklist (shipped on time).
• Three integration connectors (Salesforce, HubSpot shipped; Segment slipped).
• 10 onboarding interviews (completed; fed the checklist redesign).

Context the team added:

• The PM wants to mark KR2 as "done for the platforms that mattered" and move on.
• A stakeholder is asking whether the team "hit its numbers" for a leadership readout, and there is

pressure to present the quarter as a win.

• Evidence quality varies: activation is well-instrumented; conversion attribution is weaker; the

support-ticket rise has not been root-caused.

Audience for the artifact: the squad, their group PM, and a leadership readout. The team needs an honest score, evidence quality, learning, and next-cycle recommendations.

<!-- PM-Skills | https://github.com/product-on-purpose/pm-skills | Apache 2.0 -->

Sample: measure-okr-grader. Storevine Campaigns Q3 2026 Cycle Review

Scenario

Storevine's Campaigns team is closing the Q3 2026 cycle. The OKR set was authored in late June using foundation-okr-writer (see the corresponding writer sample at library/skill-output-samples/foundation-okr-writer/sample_foundation-okr-writer_storevine_campaigns-q3.md). The cycle ended September 30. Final values are now in for KR1 and KR3; KR2's 90-day cohorts are partially complete (the 60-day intermediate is available, the 90-day final is not yet observable).

The team wants a cycle review they can take to the Q4 planning workshop. The growth-pm runs measure-okr-grader with the original OKR set, the final and interim KR values, the cycle's narrative, and the initiative status.

The cycle had a mixed result. KR1 hit hard. KR2 trended below projection. KR3 guardrail held. Initiative 2 (Templates v2) underperformed expectations and the team needs to decide whether to retire the thesis or carry it.

Source Notes:

• Storevine is fictional
• All metrics [fictional]
• Pairs with library/skill-output-samples/foundation-okr-writer/sample_foundation-okr-writer_storevine_campaigns-q3.md
• Aspirational OKR scoring follows the Google convention (0.6 to 0.7 sweet spot for aspirational)
• Committed and compliance_or_safety scoring conventions are not exercised in this sample; see the workbench thread for committed and compliance_or_safety scoring examples

Prompt

measure-okr-grader

Original OKR: see sample_foundation-okr-writer_storevine_campaigns-q3.md
Cycle: Q3 2026 (July 1 to September 30, 2026)
OKR type: aspirational

Final KR values:
- KR1 (weekly active senders): 26% [fictional] (target was 28%, baseline 14%)
- KR2 (90-day campaign retention): 60-day cohort interim is 19% [fictional];
  full 90-day target was 38% (baseline 22.8%); 90-day final not yet observable
- KR3 (guardrail, median CTR): 3.6% [fictional] (target was hold at or above
  3.4%, baseline 3.4%)

Guardrails:
- Unsubscribe rate ended cycle at 0.81% [fictional] (baseline 0.72%, threshold 0.95%)
- Spam complaint rate ended at 0.05% [fictional] (baseline 0.04%, threshold 0.1%)

Initiative status:
- Initiative 1 (Lifecycle email triggers): shipped July 22 as planned. Open
  rate on triggered emails 38% [fictional]. Click-through to send action
  21% [fictional].
- Initiative 2 (Templates v2): shipped 8 of 12 templates by Sept 30
  [fictional]. Adoption among first-time repeat senders was 31% [fictional],
  below the 60% [fictional] target. Templates that did get adopted showed
  no significant retention difference vs. generic templates (Q3 cohort).
- Initiative 3 (Campaign performance feedback loop): shipped Aug 28.
  Adoption among repeat senders 64% [fictional]. Self-reported usefulness
  4.1 / 5 [fictional].

Cycle context:
- Onboarding team's Q3 KR (boost Campaigns activation rate to 50%)
  hit 47% [fictional]; close enough that we did not rebaseline KR1
  mid-cycle.
- Mid-cycle, the team caught a bug in lifecycle trigger eligibility logic
  (week 3 of August). Fixed in 4 days but ~120 merchants [fictional]
  received duplicate triggers. No measurable impact on unsub rate.

Stakeholder interpretation: Campaigns leadership reads the cycle as a
clear win on activation-to-recurring-send (KR1) and a learning result
on retention-thesis (KR2 + Templates v2). Looking for grader's
independent read.

Output

# OKR Cycle Review: Storevine Campaigns Team, Q3 2026

## Summary

- Scope: Campaigns team
- Cycle: 2026-Q3 (July 1 to September 30, 2026)
- Overall interpretation: strong aspirational outcome on KR1, partial signal
  on KR2, guardrail held. The KR2 90-day window extends past cycle close, so
  a single objective-level number is misleading. The honest read is that the
  team validated the engagement thesis (KR1) and invalidated the
  templates-as-retention-driver thesis (Initiative 2).
- Evidence confidence: high for KR1 and KR3, medium for KR2 (interim only),
  high for guardrails.
- Main learning: lifecycle triggers drove the engagement lift; templates v2
  did not move retention. The Campaigns engagement-causes-retention belief
  held at the leading-indicator level (KR1) but the lagging-indicator
  evidence is not yet in.

## Scorecard

- Objective: Make Campaigns the marketing tool merchants reach for every
  week.
  - Rough qualitative read: validated lift in weekly engagement; retention
    causality still open. Do not collapse this to a single average; KR2
    is not yet observable at 90 days.

- KR1: Increase weekly active senders from 14% to 28% [fictional] of
  Campaigns-activated merchants by 2026-09-30.
  - Actual: 26% [fictional] (as_of 2026-09-30, trailing 4-week average).
  - Score: 0.86 (aspirational). Calculation:
    (26 - 14) / (28 - 14) = 12 / 14 = 0.857.
  - Evidence confidence: high. Amplitude `campaign_sent` event well-
    instrumented; cohort definition stable across cycle (Onboarding team's
    near-miss kept the activated-merchant denominator within tolerance).
  - Interpretation: lifecycle triggers (Initiative 1) materially moved the
    metric. The triggered-email-to-send conversion (21%) translates cleanly
    into the weekly cohort lift.

- KR2: Increase 90-day campaign retention from 22.8% to 38% [fictional] for
  merchants who send their first campaign in Q3.
  - Actual: not-yet-observable. Q3 cohorts' 90-day windows extend into late
    December. Interim 60-day signal is 19% [fictional].
  - Score: deferred. Final score available 2026-12-31. Best-estimate
    projection if forced today is 0.10 to 0.20 (aspirational), well below
    the 0.6 to 0.7 sweet spot. See Risks in Interpretation below; do not
    treat the projection as the final score.
  - Evidence confidence: medium. The 60-day interim is reliable; the 90-day
    projection assumes Q2 cohort decay shape, which may not apply to a
    cohort with different engagement composition.
  - Interpretation: retention thesis is trending below projection. Hold the
    final score until 2026-12-31 and grade then.

- KR3 (operational_health; indicator class guardrail): Hold median
  campaign click-through rate at or above 3.4% [fictional] across all
  Q3 sends.
  - Actual: 3.6% [fictional].
  - Score: pass (operational_health; threshold held within band).
    Improved by 0.2 percentage points above the baseline.
  - Evidence confidence: high.
  - Interpretation: lifecycle triggers did not degrade send-quality. This
    is meaningful. The most common failure mode for "send more" initiatives
    is engagement collapse; the team avoided it. Per the indicator-class
    `guardrail` rule, KR3 is reported as its own signal and is NOT averaged
    into the primary objective score.

## Objective Interpretation

- Result: aspirational success on activation engagement (KR1); aspirational
  shortfall (likely) on the retention thesis (KR2, score deferred). The
  guardrail held.
- Why: Initiative 1 (lifecycle triggers) was the load-bearing bet for KR1
  and it worked roughly as hypothesized. Initiative 2 (Templates v2) was
  the load-bearing bet for KR2 and it under-shipped (8 of 12) and
  under-adopted (31% vs 60% target). Even templates that were adopted did
  not show a retention effect.
- What changed during the cycle: mid-cycle bug in lifecycle eligibility
  (4 days, no measurable impact). No external surprises. Onboarding team's
  near-miss on its own KR did not destabilize our KR1 denominator.
- What the score does and does not mean:
  - Does mean: lifecycle triggers are a validated engagement lever and
    should continue at full scale.
  - Does mean: the templates-as-retention-driver thesis is weak. Adopted
    templates showed no retention difference vs generic. This is
    invalidating evidence, not just inconclusive evidence.
  - Does NOT mean: engagement does not drive retention. The KR1 lift is
    too recent to have produced measurable retention impact yet.
  - Does NOT mean: Campaigns is a finished product. KR2 remains an open
    thesis that needs a different bet next cycle.

## Evidence Quality

- Issue 1: KR2's 90-day window extends past the cycle boundary. Q3 grade
  is a projection until late December.
  - Impact: cannot give a confident final score this cycle. Forcing a
    numeric score now would mislead readers.
  - Recommended fix: in next cycle's `foundation-okr-writer` run, either
    (a) reframe KR2 as a 30-day or 60-day intermediate metric, or
    (b) accept that retention KRs are graded one quarter in arrears and
    document the convention explicitly. Choosing (b) is cleaner if the
    team plans to keep retention as a primary KR.

- Issue 2: KR2's underperformance correlates with Initiative 2's
  under-shipping (8 of 12 templates) and low adoption (31%). This confounds
  the templates-as-retention-driver test: the team did not run the bet at
  full scale.
  - Impact: the invalidating evidence is partial. A team could argue
    "we did not really try."
  - Recommended fix: do not relitigate Templates v2 next cycle without a
    clearer hypothesis about why adoption was low. The grader's read is
    that the thesis is weak enough to retire; the team should explicitly
    decide rather than carry an ambiguous bet forward.

- Issue 3: Mid-cycle lifecycle-trigger bug affected ~120 merchants with
  duplicate sends. No measurable unsub or CTR impact, but the bug is
  unrecorded in standard dashboards.
  - Impact: low for this cycle. Documented here for the cycle log.
  - Recommended fix: add a `triggered_email_dedup_failures` event so future
    cycles can detect this class of bug in real time.

## Initiative Review

- Initiative 1 (Lifecycle email triggers):
  - Linked to: KR1 primarily, KR2 secondarily.
  - Status: shipped on time (July 22).
  - Apparent contribution: high. Triggered emails reached 38% open rate
    and 21% click-through-to-send rate, translating into the KR1 weekly
    cohort lift.
  - Recommendation: continue at full scale next cycle.

- Initiative 2 (Templates v2):
  - Linked to: KR1 and KR2.
  - Status: partially shipped (8 of 12 templates).
  - Apparent contribution: low. Adopted templates showed no retention
    effect in the Q3 cohort. Even at full ship, the underlying thesis is
    not supported by the partial evidence.
  - Recommendation: retire the current framing. If the team wants to
    revisit, run `define-hypothesis` first to sharpen the sub-thesis
    (which segment, which template type, which trigger), and validate via
    `measure-experiment-design` before baking into a KR.

- Initiative 3 (Campaign performance feedback loop):
  - Linked to: KR2 primarily.
  - Status: shipped late (August 28 vs target of mid-August).
  - Apparent contribution: unclear. 64% adoption among repeat senders and
    4.1 / 5 self-reported usefulness suggest merchant demand, but the
    contribution to KR2 cannot be isolated from Initiative 1's effects.
  - Recommendation: continue next cycle to gather more data; consider
    promoting from supporting bet to candidate primary initiative if the
    Q4 retention cohort shows a feedback-loop effect.

## Learning

- Validated assumptions:
  - Lifecycle triggers materially increase weekly engagement.
  - Engagement gains do not require sacrificing send quality (KR3 held).
  - Empowered-team initiative ownership produced a learning-grade Q3, not
    just a delivery-grade Q3.

- Invalidated assumptions:
  - Templates v2 as the primary retention-driver lever. The adopted cohort
    showed no retention effect. This is the strongest invalidating signal
    of the cycle.
  - Designer capacity to ship 12 seasonal templates in Q3 was overestimated;
    8 of 12 was the actual capacity. Revise next cycle's planning.

- Surprises:
  - Initiative 3 (feedback loop) shipped late but adopted high. Adoption
    rate suggests merchant demand for post-send analytics is stronger than
    the team expected. Worth a deeper investigation.
  - The 60-day interim for KR2 came in lower than the Q2 baseline cohort
    despite KR1 success. If engagement causally drives retention, the team
    should have seen at least a small interim lift. The flat result is
    itself information.

- Decision implications:
  - Continue Initiative 1 at full scale next cycle.
  - Retire Initiative 2's current framing.
  - Promote Initiative 3 from supporting bet to candidate primary initiative
    for KR2 next cycle.
  - Reframe KR2 measurement boundary (Issue 1) before Q4 OKR drafting.

## Next-cycle Recommendations

1. Continue lifecycle triggers as a primary lever. Set Q4 KR1 target based
   on Q3's 26% landing point, not Q3's pre-cycle 14% baseline.

2. Retire the Templates v2 thesis as currently framed. Do not re-run the
   bet without sharpening the sub-thesis first.

3. Reframe KR2 to either 60-day retention (gradeable within cycle) or
   90-day retention (graded one quarter in arrears). The former gives
   clearer cycle accountability; the latter is methodologically truer to
   the underlying behavior. Pre-decide before Q4 OKR drafting.

4. Investigate the engagement-without-retention finding. If KR1 lifts
   continue without moving retention, the engagement-causes-retention
   belief is wrong and the team's strategy needs to change.

5. Hand-off:
   - `iterate-lessons-log` for the Templates v2 invalidation (organizational
     memory).
   - `iterate-retrospective` for the team-process learning on designer capacity
     estimation.
   - `define-hypothesis` for "Campaigns engagement causally affects 12-month
     merchant retention" with a defined invalidation criterion.
   - `measure-instrumentation-spec` for the `triggered_email_dedup_failures`
     event.
   - `foundation-okr-writer` for next-cycle drafting once KR2 measurement boundary
     is resolved.

## Risks in Interpretation

- A naive 0.62 rough-average objective score would include a forced KR2
  score that is not yet observable. Reading 0.62 as "the team hit the
  aspirational sweet spot" would be misleading. The honest read is "0.86
  on KR1, retention thesis still open at 90 days, guardrail held." Avoid
  collapsing heterogeneous KR types into a single number.

- Initiative 2's under-shipping confounds the Templates v2 invalidation. A
  reasonable counter-read is "we did not really test it." The grader's
  stronger evidence is the no-retention effect among the 31% who did adopt;
  that is the part that says the thesis is weak even at full adoption.

- Stakeholder framing of "clear win on activation, learning on retention"
  is broadly correct but understates the invalidating signal on Templates
  v2. The Q4 planning workshop should explicitly decide whether to retire
  or rework the thesis rather than leave it as ambiguously "ongoing."

- KR3 (guardrail) holding is good news but is not by itself proof of
  safety. Two cycles of held guardrails would strengthen the case that
  lifecycle triggers do not degrade send quality at scale.

## Source of Truth

go/okrs-q3-2026-campaigns (Confluence). This artifact is a review document,
not the canonical OKR record.

OKR Cycle Review: {team or scope name}, {cycle}

Output template for the measure-okr-grader skill. Section order is canonical and enforced by the skill's Output Contract. Remove this guidance blockquote and all template comments before finalizing the artifact for the user.

Summary

One-paragraph orientation: scope, cycle, overall interpretation (aspirational sweet spot? clear win? committed miss? mixed?), evidence confidence overall, and the main learning the cycle produced. Do NOT lead with a single number; OKRs with heterogeneous types do not average meaningfully.

• Scope: {team | product-area | department | company}
• Cycle: {2026-Q3 | 2026-H2 | 2026 | launch window | custom}
• Overall interpretation: {qualitative read; do not force a single average across mixed KR types}
• Evidence confidence: {high | medium | low | mixed}
• Main learning: {one sentence on the most load-bearing learning}

Scorecard

Each KR is scored using the convention for its OKR type from the canonical 5-value enum: committed | aspirational | learning | operational_health | compliance_or_safety. Indicator class (leading | lagging | guardrail | health | evidence_generation) is independent and applies on top of the type. Numeric scores belong to aspirational KRs only; committed KRs are pass or fail; compliance_or_safety KRs are binary (no partial credit, no retroactive scope shrinkage); operational_health KRs are pass | fail | drift-within-tolerance; learning KRs use validated or invalidated language. Special states: not-yet-observable for cycle-window extensions past close; not-yet-fully-observable for committed or compliance_or_safety KRs with partial coverage. KRs with indicator class guardrail are surfaced as their own signal and never averaged into the primary objective score, regardless of OKR type.

• Objective: {original objective text}
• Rough qualitative read: {one-line summary; do NOT force a single numeric average across heterogeneous types}

• KR1 ({OKR type}; indicator class {indicator class}): {original KR text, baseline to target}
• Actual: {value, with as_of date; not-yet-observable if cycle window extends past close; not-yet-fully-observable if committed or compliance_or_safety with partial coverage}
• Score: {numeric on 0-1 scale for aspirational; pass | fail for committed; binary met | not-met for compliance_or_safety; pass | fail | drift-within-tolerance for operational_health; validated | invalidated for learning; deferred for not-yet-(fully-)observable}
• Evidence confidence: {high | medium | low | unknown}
• Interpretation: {what this score does and does not mean; if indicator class is guardrail, note that the score is reported separately and not averaged into the primary score}

• KR2: {as above}

• KR3: {as above}

Objective Interpretation

Synthesize a qualitative read of the objective. Avoid naive averaging when KRs have different types. State explicitly what the score does and does not mean so future readers cannot over-read it.

• Result: {qualitative summary}
• Why: {what drove the result; which initiatives carried the load}
• What changed during the cycle: {scope shifts, market changes, team changes, dependency shifts}
• What the score does and does not mean:
• Does mean: {1 to 2 statements}
• Does NOT mean: {1 to 2 statements that prevent over-reading}

Evidence Quality

For each significant evidence issue, name the issue, its impact on the score, and a recommended fix for next cycle. Do not paper over weak evidence with precise numbers.

• Issue 1: {description}
• Impact: {how this affects the score's reliability}
• Recommended fix: {next-cycle measurement change}

• Issue 2: {as above}

Initiative Review

For each initiative the team ran, name which KR it was expected to move, whether it shipped, what its apparent contribution was, and whether the evidence supports continuing, retiring, or reworking it. Separate ship-status from KR-impact.

• Initiative 1: {name}
• Linked to: {KR1, KR2, etc.}
• Status: {shipped on time | shipped late | partially shipped | not shipped}
• Apparent contribution: {high | medium | low | unclear}
• Recommendation: {continue | retire | rework with sharper hypothesis}

• Initiative 2: {as above}

Learning

Distinguish customer or product learnings (carry forward to next cycle), team-process learnings (hand to retrospective), and measurement learnings (hand to instrumentation or dashboard skills).

• Validated assumptions: {list}
• Invalidated assumptions: {list}
• Surprises: {findings the team did not anticipate}
• Decision implications: {what the team should do differently next cycle}

Next-cycle Recommendations

Numbered list. Each recommendation either drives next-cycle OKR drafting or hands off to a specific downstream skill. The grader's job is to set up the next cycle, not to write its OKRs.

1. {recommendation} 2. {recommendation} 3. {recommendation} 4. Hand-off:

• iterate-lessons-log for {what learning needs organizational memory}
• iterate-retrospective for {what team-process work needs reflection}
• define-hypothesis for {what assumption needs an explicit test}
• measure-instrumentation-spec or measure-dashboard-requirements for {what measurement gap needs filling}
• foundation-okr-writer for {next-cycle drafting note}

Risks in Interpretation

Make explicit any places the score could mislead a reader. The grader's job is to protect the integrity of the OKR operating system, not to manufacture certainty.

• {risk 1}
• {risk 2}

Source of Truth

{URL or path to the live OKR tracker; this artifact is a review document, not the canonical record}